The Atlas humanoid robot crosses the stage during a Hyundai Motor and Boston Dynamics press conference ahead of CES 2026 in Las Vegas on Monday (local time). AP-Yonhap
LAS VEGAS—Hyundai Motor Group announced Thursday (local time) that it has completed the development of its long-awaited artificial intelligence (AI) chip for robots, taking an important step toward building a robust supply chain and infrastructure in the era of physical AI.
The global automaker and emerging robotics giant has partnered with DEEPX for three years to jointly develop on-device AI chips. DEEPX is an advanced AI chip manufacturer and computing solutions developer.
The group's Robotics LAB showcased the milestone at CES Foundry 2026, the tech fair's first subprogram dedicated to discussions about AI, quantum computing and other cutting-edge future technologies.
On-device AI chips are also ready for mass production, helping automakers build a stable supply chain for the entire robotics business.
During CES 2026, Hyundai Motor Group highlighted physical AI, the integration of hardware and artificial intelligence, as a key revenue area for the future. The automaker outlined its vision to deploy several advanced robots, including Atlas and the multi-purpose self-driving Mobile Eccentric Droid, and leverage them into smarter, more efficient workplaces.
Hyun Dong-jin, vice president of the group and head of the Robotics LAB, said, “In order to realize physical AI, Hyundai Motor Group is developing AI solutions and related software with the vision of robotizing space.'' “Our ultimate goal is not just robot development per se, but building a sustainable robot ecosystem,” he said.
In an aging society, many countries are facing social issues such as labor shortages. According to the automaker, this will gradually increase the demand for robots, especially in the workplace.
Hyundai Motors and Kia Motors expect their latest on-device AI chips to help build a stronger and more stable physical AI infrastructure across key value chains, from vehicle production to logistics.
The chip operates on ultra-low power and enables robots to perform human-like tasks such as real-time perception and decision-making. According to the company, the AI semiconductor is particularly useful in environments with unstable network connections, such as underground parking lots and distribution centers.
The chip does not rely on cloud servers to operate, making it more responsive and less exposed to potential hacking threats.
The company said the latest breakthrough will enable the group to expand its revenue stream to airports, hospitals and other work sites, significantly increasing demand for integrated AI robotics solutions.
